R-Verter with pass-thru

[rickcollette]

Does anyone have an idea on how to build this with an SIO passthru?
Or is there an existing discussion on this, some other resource, etc?

 

 

[mozzwald]

5 minutes ago, rickcollette said:

Does anyone have an idea on how to build this with an SIO passthru?

All you would need to do is connect the 13 SIO input pins to another port and connect the pins on the left side of the schematic to their respective SIO line. Basically 2 SIO ports connected together and tap the signals needed for your Rverter

[rickcollette]

I did not realize it was that easy.  I was expecting some kind of filtering or something to keep potential crosstalk down.

Thank you @mozzwald !

 

[_The Doctor__]

now if we could just get the drivers and programs to realize that rts is cross connected to cts on modems and other devices... and to toggle it off when the sio or atari is too busy doing other things to receive the data... and back on when it's good to send again we'll be in business... an issue that seems to be confused at times. Generally cts isn't an issue coming from the Atari so many rverters only put out rts as most devices are never fed data quick enough to overflow their buffers... but that could be an issue if we send to super slow devices with little or no buffers. Though there was some rverter that provided that signal as well

[mozzwald]

46 minutes ago, rickcollette said:

I did not realize it was that easy

Depending on what you need the Rverter for you might be interested in FujiNet which supports modem emulation (among other things  ) if that's what you are looking for.

 

Also, I have a pass through breakout that may be useful for connecting to your Rverter 

 

https://fujinet.online/shop/hardware/sio-pass-through-breakout-board-v2-2/

[rickcollette]

It is for a modem device - but I'm trying to learn this stuff ; so im eeking my way through, bit by bit..

 

[Tillek]

2 hours ago, rickcollette said:

It is for a modem device - but I'm trying to learn this stuff ; so im eeking my way through, bit by bit..

 

I think he's more clarifying whether or not you've heard of the FN.

 

[Tillek]

One concern would likely be the command signal and the logic to ensure it only responds to the actual R: device call, especially if there is a pass through, that would imply you're likely to have other devices on the chain.  I think this was one of the drawbacks to the "poor man's  R:Verter" that the SysOp of Cheeze Daddy's was using which I think was pretty much this setup.

 

Some other topics on this subject, many with pictures.  Some with many pictures.

 

 

[rickcollette]

@Tillek Awesome! thank you!

 

[Tillek]

Too bad you weren't on the Atari BBS Chat tonight.... Someone is working on a plan to put an R-Verter II and a WiModem 232 in a 1030 with the LED working.

 

We're having that discussion as we speak (with screen share on his schematics).

 

[+bf2k+]

Yep... it was an interesting Zoom meeting.

[rickcollette]

nice! Thats actually what i am doing (attempting to do) - something close to that... 

I have a few boards made now that are just rs232->esp w/wifi.

The next version is with a builtin rverter type thing - what clearly needs some hammering on to work properly.

 

[manterola]

20 hours ago, Tillek said:

Too bad you weren't on the Atari BBS Chat tonight.... Someone is working on a plan to put an R-Verter II and a WiModem 232 in a 1030 with the LED working.

 

We're having that discussion as we speak (with screen share on his schematics).

 

I did something like that.. I'll find the info and post it here. But it is basically using an esp32 for TX/RX and an additional output of esp32 to turn the led on.

Edited December 4, 2021 by manterola

[Tillek]

34 minutes ago, manterola said:

I did something like that.. I'll find the info and post it here. But it is basically using an esp32 for TX/RX and an additional output of esp32 to turn the led on.

He was going a slightly different route, but still... we'd love to see what you have worked on as well.  You actually got it working?

 

I did record the chat (it was pretty big and went pretty long) so instead of just posting it, I think I'm going to chop it up into the relevant parts.  If/when I do that, if he's ok with it I'll post it here.

 

[manterola]

This is what I found:

First, the es8266 inside the Atari 1030.

 

Second, the circuit to turn the led on.

 

[Fierodoug5]

I put a fujinet in a 1030 and hooked up the original 1030 Led's to work.

 

[Jeffrey Worley]

On 12/3/2021 at 7:07 PM, Tillek said:

One concern would likely be the command signal and the logic to ensure it only responds to the actual R: device call, especially if there is a pass through, that would imply you're likely to have other devices on the chain.  I think this was one of the drawbacks to the "poor man's  R:Verter" that the SysOp of Cheeze Daddy's was using which I think was pretty much this setup.

 

Some other topics on this subject, many with pictures.  Some with many pictures.

 

 

I'm really for this idea.  I bought an original Rverter in 1984 to output to a serendipitously aquired Epson Ep22? thermal typewriter.  The serial printer was not ideal for the Atari, but it did work.  In recent years I've made a number of cheap Rverter-compaitble devices but mine lack any sort of flow control whatsoever.  My point is that I think there is a real need for a proper Rverter, one that uses Command to make the device smart enough to co-exist with other devices on the chain.  Mine answers for EVERYONE.   Works fine with PBI for storage, but is obviously a fail for anything that uses SIO, like Fujinet.

[Kyle22]

If your R:Verter device is wired properly to the SIO Proceed and Interrupt lines, then it can do flow control if your R:Handler is written correctly.

 

Look how FujiNet works.

 

[Jeffrey Worley]

I know.  My cheapo solution is kinda crummy but it does work.  Should you need a better behaved device, the real-deal is it.  Fujinet loads the serial handler, actually emulates a Hayes-Compatible modem, and supplies a basically perfect link from Atari RS232 (Bobterm or what I originally started with SMARTTERm, on cartridge.  Boy howdy, that last made me get a disk drive.

 

Jeff

[manterola]

Like many of these projects there is a software component and a hardware component, and obviously both need to converge.

That's the beauty of Fujinet, you have all the hardware, but also the software running in the esp32, and also the configurator which is the Atari software that is autoloaded to the Atari computer by the Fujinet device (and a lot more stuff).

Now the rverter way, is a lot simpler underneath, but more complex from the pure user perspective. It uses the rverter handler in the Atari, that you need to load previously using DOS (load manually, but can be automated).

Then you also need the rverter circuit that basically uses a 74ls00 to disconnect TX and RX from DIn and DOut of the SIO bus: if MotorCtl signal is not enabled the TX and RX are disconnected.

Finally, after the rverter circuit you have your "Modem" device .  That device use to be a real modem to connect to the telephone line, nowadays is usually some modern microprocessor system like esp32 or similar running a Modem emulator firmware.

There are multiple firmwares for modem emulation if you look in GitHub.  for example: zimodem, espmodem or even the default firmware from the esp8266.

Okay, long enough post, but I would like to explain the circuit to turn the led on in the Atari 1030 modem, before ending this.

In the original circuit, the ready led is connected to the Data Carrier Detect pin of the 339 IC. When carrier is detected that pin goes to Gnd so we have current from 5V passing thru the resistor and LED, so led is on. So I just tried to replicate the same idea using a pn2222 transistor. In my case, I used the ZiModem firmware for esp8266, which uses gpio02 to indicate that a telnet connection has been established. So that gpio02 is connected to the transistor to make it turn on and turn off, which provide a parallel path for the current to flow, and turn the LED on, while keeping the original functionality of the 1030 modem intact.

I hope this may help others implement something similar and have some fun in the process.

 

 

 

 

Edited December 5, 2021 by manterola

[clh333]

To confirm (playing catch-up here): 

 

The original "R-verter" was a product of the 1980s designed to obviate the need for an 850 interface, as described here: https://www.atarimagazines.com/compute/issue51/237_2_NEWS_PRODUCTS_RS-232_Modem_Adapter_For_Atari.php.  The term is now generic, as "band-aid" or "coke", meaning principally a signal-shifting device translating TTL to +12VDC/-12VDC (ideally).

 

An "R handler" is a device driver that replaces or functions in lieu of the R: driver the 850 downloads on console startup, provided the 850 is already started when the console starts, as the console OS does not have an R: CIO routine built in (unlike P: and E:...).

 

A "1030" is an Atari 300-baud modem having two SIO ports and no "D-sub" 9- or 25-pin ports for direct RS-232-level connection as does the SX212.

 

The ESP8266 is, according to Wikipedia, "The ESP8266 is a low-cost Wi-Fi microchip, with built-in TCP/IP networking software, and microcontroller capability, produced by Espressif Systems in Shanghai, China"

 

Is so?

 

-CH-

[Tillek]

Sounds about right/close enough (I'm sure some would argue for hours over minutia that aren't really relevant).